Jeff King <p...@peff.net> writes: > But for git-fetch, our primary purpose is receiving data and writing it > to disk. We should be checking all of our write()s already, and SIGPIPE > is just confusing.
Yup, sounds like a very sensible argument. > So I'd actually be fine with just declaring that certain commands (like > fetch) just ignore SIGPIPE entirely. Me too (to me, "actually be fine with" would mean "that's a larger hammer alternative, which is OK, while the base solution is fine, too"). Thanks.
